Psychiatric medications have been established as an efficacious treatment for pediatric anxiety and depression, with approximately 60% of youth responding to pharmacotherapy alone and over 75% responding to multimodal intervention (March et al., 2004, Walkup et al., 2008). However, recent research has indicated that medication adherence for these conditions is suboptimal, with over 50% of children discontinuing treatment before the period recommended for full therapeutic benefit (Murray et al., 2004; Richardson et al., 2004). This is highly troubling given that these conditions are relatively common and highly debilitating psychiatric conditions, which are associated with substantial functional impairment, reduced quality of life, and suicidality (Langley et al, 2004, Puig-Antich et al., 1993; US Surgeon General, 1999). Moreover, poor adherence may have a widespread effect on treatment effectiveness for these conditions, as an estimated 2.6% of American children receive antidepressant medication in a given year (Olfson & Marcus, 2009). Unfortunately, relatively few empirical data exist about variables that may relate to medication adherence in pediatric anxiety and depression. With this in mind, the goal of the proposed investigation is to examine potential mechanisms of such adherence, with a particular focus on the therapeutic alliance (Shirk & Karver, 2003). Indeed, children and parents who do not feel a bond with the clinician, agree with the clinician on tasks to be performed in treatment, or have unrealistic goals for outcome may be less likely to continue pharmacotherapy in the face of side effects and inconvenience. Adolescent psychiatric outpatients (N=150) as well as their parents and clinicians will be evaluated through self-reports and tracking of pharmacy records to evaluate the role of therapeutic alliance and other behavioral mechanisms of medication adherence. Additionally, this fellowship will provide training to the applicant to prepare him to become an independent, externally funded investigator through a series of personalized activities (including additional coursework, didactic activities, and direct mentoring and consultation). Potential implications of this research are multifaceted. First, future research would need to be focused on developing evidence-based interventions to improve alliance (e.g., communications skills training for the working alliance; Meystre et al., 2012). Second, training curricula of medical personnel (e.g., psychiatry residents) would be indicated to include instruction in developing a therapeutic alliance. Third, identifying these mechanisms could assist in creating an idiographic approach to patient care, where specific interventions to improve medication adherence are employed only for patients who may need them. Thus, the proposed research has wide-ranging implications for future research and the training of medical personnel.